According to Techopedia, green computing is ‘’The environmentally responsible and eco-friendly use of computers and their resources.” Even though that says a lot, it also does not say everything. Green computing, and similar terms, have been around for at least 2 decades since the launch of the EPA but what does it really mean in today’s world? Techopedia presents four approaches:
Green computing is thus not just that you are putting your computer in sleep mode after some time of being idle; it comprises the entire lifecycle of all relevant production as well as usage of IT infrastructure, from data centers to servers to network equipment.
When we look at the data it is clear why green computing is important: The Information and communication technology (ICT) sector is responsible for between 1.8% and 2.8% of global greenhouse gas emissions according to a report by the Association for Computing Machinery and according to the International Energy Agency (IEA), data centers and data transmission networks each account for 1% of global electricity consumption in 2019. As a provider of IT infrastructure, we take our part in slowing down climate change seriously by implementing green energy strategies in our entire IT infrastructure footprint.
Aside from the clear benefit of doing your part in slowing down climate change, there are also very tangible advantages you can benefit from by taking green computing seriously:
How tangible are carbon emissions for you today? Aside from the effect on climate change, due to regulations from various governments, carbon emissions have become a tangible concern for businesses today — either for obtaining or keeping certain certifications, but also the demand from governments to report and lower your carbon emissions.
Green computing does not have to cost you money, it can help you save money, especially today with energy prices increasing and your IT infrastructure running 24/7. Achieving this can be done by writing overly complex software, but it can also be quite easy by adjusting settings from the manufacturer of your equipment.
Purchasing hardware which lasts longer may by its design increase the efficiency of your hardware and will take less manual time to replace. In return, you will have more resources available for other upgrades. Ultimately, if it is produced to be more carbon neutral and operates longer, it indirectly decreases the need for mining several raw materials too. Building this way helps use your power availability: in these days of rising power prices, this is especially beneficial for power efficiency. Another advantage is that you are delaying your CapEx (Capital Expenditures).
Disadvantages of green computing
There are no disadvantages, right? Well, the answer is that it depends how we look at things. If you really zoom in on specific parts of the entire chain of choices, there might be. For example, in today’s world, purchasing carbon-neutral products usually comes with a price tag due to the higher cost of production (or even the cost of the marketing of it.). Zooming in on that, a higher price tag may be seen as a disadvantage, however, if we look at the specifications of various hardware it may be that the carbon-neutral product utilizes less energy than a cheaper product. Looking at the entire chain of choices and their lifecycle is crucial to determining its disadvantage.
A second disadvantage may be that if you are increasing the use of specific equipment, you may not always be using the latest and greatest, which for 99% of applications and use cases is fine, but for some innovative technologies, this may not be sufficient.
Most of the IT infrastructure used by organizations is in data centers, therefore your data center selection, or the provider’s data center choice is a crucial one if you are serious about going green for your compute needs. So, what are key factors you need to look for in data centers, or that data centers can utilize to reduce their carbon footprint?
One of the easiest things is the way a data center is cooled. If this is done in the traditional way with compressors, it is very power heavy. While this was the standard 10–15 years ago, there are still plenty of data centers operating on this model today.
Second, does the power the data center use green energy?
Another is the design of the data center, is it designed to leverage its residual-heat? Where does the heat go and how is it processed? Most data centers still have big open floors with cooling and the heat would just go somewhere else. We see a shift where these are more contained with cold and hot corridors and a management system to make the best of both.
At the time of writing, in our data center we improved our power usage effectiveness (PUE) from 1.35 to 1.15. We also reduced our fuel consumption by almost 15 percent (489 liters out of a total of 3302). All of this combines to reduce our own data center’s footprint by over 13 percent, dropping from 436 tons of CO2 eq to 378 tons CO2 eq. Read more about our heat pump project on the Smartdc website.
One of the most well-known measures in the industry is PUE. It tells you how much energy you need to operate your data center and its waste. For example, if you have a PUE of 1.3, this means that for every ampere (amp) your IT infrastructure requires 0.3 amps to operate on a functional level — we’re talking about the energy that’s being used for cooling, lights etc. The closer you get to 1, the less power you need to operate, the lower your carbon emission and the higher your sustainability rate is.
Implementing a green computing strategy is challenging, especially as you need to look at the entire chain, and if you are in an organization just leveraging compute resources, you may not always have a direct say in how your providers manage the supply chain and lifecycle. However, this shouldn’t be the case, if you ask questions from your supplier about how energy efficient their operations are, and how their supply chain is optimized, it can change your supplier’s way of thinking and operating, as the goal of the providers is to empower you.
That said, there are basically two options you can go with if you want to make changes to the infrastructure: Gray Field Migration and a Green Field Rollout. Whereas Gray Field is replacing and fixing what’s already there and Green Field is rolling it out with everything you build from scratch. Obviously from a technical and operational standpoint, a Green Field Rollout should be the easiest as you start from the beginning, whereas in a Gray Field Migration you’d have to do a proper migration from a to b with all the challenges that come with this without taking green compute into account in the first place.
As there is plenty of information available on the web about how to do a Gray Field Migration from a technical and operational standpoint, and assuming you and your stakeholders are convinced of a Green Compute Strategy, we’ll leave that for what it is. Yet a Green Field Rollout may be very difficult; there may be a higher dollar value attached to convincing your stakeholders of a Green Compute Strategy instead, as at first glance, there may be a higher dollar value attached to it and changes are not visible with immediate effect.
Here’s a few tips and tricks on how to convince your board or decision-making body that a Green Compute strategy is worth it:
Every single provider in the world should have metrics available for how they perform from an environmental standpoint, whether it is the PUE for a data center, power usage for servers, or what their supply chain looks like — if you find yourself planning a new IT Infrastructure rollout, ask for those metrics as at the end of the day, it will reflect on your business practices as well.
Having the data is one thing, but what does that mean for you and your organization? Show your stakeholders the positive impacts going for a Green Compute strategy will bring, whether that is a long-term savings on energy costs, a longer lifecycle of hardware or maybe even additional marketing value it brings to market your products or internal reporting on sustainability.
It may be that nothing will happen, however that strongly depends on the field of work you’re in. Are some of your customers already asking for sustainability metrics? Count on that demand increasing over the coming years. Does your current bottom line get affected by rising energy prices? Bet that with a Green Compute Strategy, you save dollars there. In other words, sustainability and the environment are hot topics in today’s world. However, everyone interprets them differently, while it is relatively simple to frame in your business’s context. Because that is from where you’ll be able to drive impact.
There are several government regulations that require companies to move to sustainable solutions and lower carbon emissions. One day, all companies need to conform to these regulations, and they will only become stricter by the year as governments are working heavily on this topic. You might be better off ahead of the game!
If you made it this far in the blog, you might have guessed it: Strategizing the entire chain and not zooming in on one specific aspect. As strategies may be different for different companies operating in various segments, we are providing insight into the green computing strategy of i3D.net here based on the four approaches as defined earlier in this post.
As i3D.net does not design or manufacture equipment itself we have merged these two pillars as the strategy is similar: Look at a multitude of variables for deciding which vendor, and which specific SKUs to integrate within our global IT infrastructure footprint.
One of the first decisions impacting energy efficiency is already made in point 1, some CPUs have a higher power draw than others which is one of the variables we utilize for our strategy in point 1.
As we operate a global IT infrastructure footprint, we made the decision to partner up with local parties to help us install and maintain the infrastructure, so we limit the miles our colleagues spend on airplanes.
Our global infrastructure footprint is utilized by a multitude of customers: where we can, we optimize settings to save energy, which made a significant impact when we first implemented this strategy. In 2022, we worked together with one of our customers when enabling power saving mode on our machines and it allowed for an average saving on power usage between 10.14% on high CPU load and 31.24% when the server was left idle, without affecting the performance of this customer.
The disposal of our infrastructure is managed very carefully. We balance the lifecycle of our servers not only by demand in the market but also by the power usage of servers. Some models we keep in our product portfolio for a longer time, others have been beaten by their successors when it comes to power consumption, making it less beneficial for the environment to keep them running. Out of everything we do dispose of, the largest part is sold off to other parties that have a demand for the equipment, something which we can do as we only purchase equipment from top-tier providers. We use the company Sims Lifecycle Services, which practices carbon-friendly policies for recycling, to recycle the fractional amount of equipment left over.
Read our upcoming blog posts for deeper dives into the topic.
Implementing a green IT strategy is rather complicated and comes with its challenges, though, it is extremely necessary for our future. By being able to identify sustainable actions in our industry, we can play a role in our sustainable future.